Statistical Analysis Of The Fifty-State Strategy
Harvard University's Elaine Kamarck has produced a paper that tries to quantify the impact of the fifty state strategy on House races in the 2006 elections, "Assessing Howard Dean's Fifty State Strategy and the 2006 Midterm Elections". You can read the paper here, and guest access is not hard to get in order to peruse the entire text. It is worth a read for anyone interested in the debate over narrow targeting and television air time versus the fifty state strategy and on the ground organizing. From page four of the paper:
I doubt that this paper will settle the question over the effectiveness of the fifty-state strategy anymore than it already has been settled, but I do believe that this paper offers some much needed factual overlays to the debate. If anyone wishes to argue that the fifty-state strategy hurt Democrats in key congressional districts, they should first have to answer to the analysis presented in this paper.
By the end of Dean's first year in office, the fifty state strategy was in full swing. The DNC was paying for 183 people working for state parties as part of their coordinated campaigns. Most of this work went on below the radar screen.
I admit I was unaware of the wide scope of the fifty-state strategy--183 paid organizers is quite a large amount. It is certainly a very expensive electoral and party-building strategy that shifts a huge amount of funds away from television advertising during the final few weeks of the campaign in selected, narrowly targeted districts. Assessing the effectiveness of this strategy with an objective eye thus becomes increasingly important, since tens of millions of campaign dollars are at stake, and both those within the Democratic party infrastructure who favor the fifty-state strategy, such as state party chairs, as well as those who oppose it, such as consultants for Democratic campaign committees, stand to either gain or lose a huge amount of money depending on the scale to which the strategy is implemented. After a detailed explanation of her methodology and why she had to focus on House races instead of statewide contests, Kamarck offers the following analysis (p. 7-8):
As Table 1 indicates, those congressional districts where the DNC had paid organizers on the ground for over a year more than doubled the Democratic vote over what would have happened due to forces outside the control of the Party, such as the war in Iraq and the unpopularity of a Republican President. This is a powerful testament to the value of a long-term party building approach. Gains in the Democratic vote occurred where the Democrat won and where the Democrat lost. The Democratic candidate won in 20 of the 39 districts where the DNC had organizers but this should not detract from the accomplishment of dramatically increasing the vote in those districts. In some places the organizer's initial and primary responsibility was to increase the vote in order to impact statewide races. In others the Democrats created a swing district where there had been none before.
This finding, while an impressive testament to the value of campaign activity in a district, does not settle the argument over the 50 state strategy. Many of the districts that had the benefit of a DNC organizer were also districts that were targeted by the Democratic Congressional Campaign Committee (DCCC) and received significant national attention. So the next step in trying to determine what matters is to arrange the districts targeted by Dean according to the amount of money the DCCC contributed to the Democrat in those districts. I did this in Appendix B, which shows how much the DCCC contributed in each of these seats and the percentage change in Democratic vote share for that district. Clearly, there is not a simple linear progression: more money is not necessarily related to greater increases in the Democratic vote.
For the sake of brevity, Table 2 summarizes the results of the analysis. It suggests that, in the absence of significant amounts of DCCC money, the presence of a DNC organizer in a congressional district puts the average Democratic increase in the vote significantly above what would be expected simply given the anti-Republican currents in the country in 2006. Districts that had less than $10,000 from the DCCC still exceeded the average national increase by nearly 3 percentage points. Districts that received between $10,000 and $100,000 exceeded the average national increase by nearly 4 percentage points. Not surprisingly, those districts that received between $100,000 and $200,000 in DCCC contributions exceeded the national average by over 8 percentage points. These districts also exceeded the average increase for the districts with a DNC organizer. Obviously money matters. But what is interesting about this table is how much can be accomplished with organization. Since there were only two districts that were targeted by the 50 state strategy that received more than $200,000 I would not make too much of this finding. Keep in mind that it is often argued that there are diminishing returns to money in politics.
While I am not convinced by the amounts of money she lists as independent expenditures by the DCCC in the districts in question (her figures seems very low), overall, the message is clear: the paid organizers in these key districts led to a substantial increase in the Democratic vote share over 2002. Some may question whether a gain of three to eight percentage points is worth the huge amounts of money the DNC spent to employ these organizers, but I think it definitely was. I believe that field organizing has much longer-term effects than television-based forms of voter contact, which will benefit Democrats in the targeted areas for many election cycles to come almost no matter who the future Democratic candidates in those districts may be. Also, in the event that the country reverts back to a state of nearly even polarization, a three to eight percent advantage in key districts could easily mean the difference between being on the right side of trifecta governance and the wrong side. After the 2004 elections, we learned that only 4.7% of the country changed its mind during the 2004 campaign, but the many small campaign advantages Republicans held outside of the airwars consistently allowed them to govern with an iron fist and a 50%+1 "majority."This finding, while an impressive testament to the value of campaign activity in a district, does not settle the argument over the 50 state strategy. Many of the districts that had the benefit of a DNC organizer were also districts that were targeted by the Democratic Congressional Campaign Committee (DCCC) and received significant national attention. So the next step in trying to determine what matters is to arrange the districts targeted by Dean according to the amount of money the DCCC contributed to the Democrat in those districts. I did this in Appendix B, which shows how much the DCCC contributed in each of these seats and the percentage change in Democratic vote share for that district. Clearly, there is not a simple linear progression: more money is not necessarily related to greater increases in the Democratic vote.
For the sake of brevity, Table 2 summarizes the results of the analysis. It suggests that, in the absence of significant amounts of DCCC money, the presence of a DNC organizer in a congressional district puts the average Democratic increase in the vote significantly above what would be expected simply given the anti-Republican currents in the country in 2006. Districts that had less than $10,000 from the DCCC still exceeded the average national increase by nearly 3 percentage points. Districts that received between $10,000 and $100,000 exceeded the average national increase by nearly 4 percentage points. Not surprisingly, those districts that received between $100,000 and $200,000 in DCCC contributions exceeded the national average by over 8 percentage points. These districts also exceeded the average increase for the districts with a DNC organizer. Obviously money matters. But what is interesting about this table is how much can be accomplished with organization. Since there were only two districts that were targeted by the 50 state strategy that received more than $200,000 I would not make too much of this finding. Keep in mind that it is often argued that there are diminishing returns to money in politics.
I doubt that this paper will settle the question over the effectiveness of the fifty-state strategy anymore than it already has been settled, but I do believe that this paper offers some much needed factual overlays to the debate. If anyone wishes to argue that the fifty-state strategy hurt Democrats in key congressional districts, they should first have to answer to the analysis presented in this paper.
|
|
|
Permalink :: 6 Comments :: Post a Comment
|
In order to post a comment, you must be logged in. If you have a member account, please log in to comment.
If not, you can make an account right here. It's quick and free.
